What evidence supports the use of pembrolizumab (Keytruda) as third-line therapy for metastatic cholangiocarcinoma after progression on gemcitabine‑cisplatin and prior pembrolizumab?

Pembrolizumab in Third-Line Metastatic Cholangiocarcinoma After Prior Pembrolizumab: Not Recommended

Pembrolizumab should NOT be used as third-line therapy in metastatic cholangiocarcinoma if the patient has already received pembrolizumab in first-line therapy (as part of gem-cis-pembrolizumab), as re-challenge with the same checkpoint inhibitor after progression is not supported by evidence and contradicts guideline recommendations.

The Critical Issue: Prior Checkpoint Inhibitor Exposure

The NCCN 2025 guidelines explicitly state that pembrolizumab is appropriate for "subsequent-line (for progressive disease and with no prior treatment with a checkpoint inhibitor)" 1. This restriction exists because:

• Checkpoint inhibitor re-challenge lacks efficacy data in cholangiocarcinoma
• Patients who progress on pembrolizumab have demonstrated resistance to PD-1 blockade
• The mechanism of action remains unchanged, making response biologically implausible

What the Guidelines Actually Recommend for Third-Line

First Priority: Molecular Testing and Targeted Therapy

The EASL 2025 guidelines emphasize that all patients with inoperable cholangiocarcinoma should undergo molecular profiling to identify actionable alterations 2. For third-line therapy after gem-cis-pembrolizumab progression, the algorithm directs toward:

1. FGFR2 inhibitors (if FGFR2 fusion/rearrangement present)
2. IDH1 inhibitors (if IDH1 mutation present)
3. BRAF/MEK inhibitors (if BRAF V600E mutation present)
4. HER2-directed therapy (if HER2 amplification/overexpression present)
5. MSI-H/dMMR-directed therapy - but this requires no prior checkpoint inhibitor 1

Second Priority: Chemotherapy

The EASL 2025 guidelines recommend FOLFOX, FOLFIRI, or NalIRI+5FU as second-line therapy 2. If these were not used in second-line, they become appropriate third-line options.

The Only Exception: MSI-H/dMMR/TMB-H Tumors WITHOUT Prior Checkpoint Inhibitor

If—and only if—the patient's tumor harbors MSI-H, dMMR, or TMB-H status AND they did NOT receive a checkpoint inhibitor previously, pembrolizumab would be appropriate 1, 3. However, your scenario explicitly states the patient received gem-cis-pembrolizumab first-line, making this exception inapplicable.

Evidence for MSI-H/dMMR (When Checkpoint-Naive)

The KEYNOTE-158 trial demonstrated in cholangiocarcinoma/biliary tract subgroups:

• ORR: 40.9% (95% CI: 20.7-63.6%)
• Median PFS: 4.2 months (95% CI: 2.1 months–not reached)
• Median OS: 19.4 months (95% CI: 6.5 months–not reached) 1

This led to FDA approval for MSI-H/dMMR solid tumors "which have progressed after prior treatment and who have no satisfactory alternative treatment options" 3. The critical caveat: this approval assumes no prior checkpoint inhibitor exposure.

TMB-H Evidence (When Checkpoint-Naive)

For TMB-H tumors (≥10 mutations/megabase), KEYNOTE-158 showed an ORR of 29% 1. However, none of the 63 biliary tract cancer patients in the TMB cohort were TMB-H 1, and only approximately 4% of advanced biliary tract cancers harbor TMB-H 1.

The Anecdotal Case Report: Not Practice-Changing

One case report 4 described an 83-year-old with MSI-H cholangiocarcinoma who responded to pembrolizumab after durvalumab (anti-PD-L1) failure. While interesting, this represents:

• A single patient (n=1)
• Different checkpoint inhibitor mechanisms (PD-L1 vs PD-1 blockade)
• Not the same scenario as pembrolizumab after pembrolizumab
• Insufficient evidence to override guideline recommendations

Practical Third-Line Algorithm

Step 1: Confirm comprehensive genomic profiling was performed

• If not done: Order immediately (FoundationOne CDx or equivalent)

Step 2: If actionable alteration identified:

• FGFR2 fusion/rearrangement → Pemigatinib, infigratinib, or futibatinib
• IDH1 mutation → Ivosidenib
• BRAF V600E → Dabrafenib + trametinib
• HER2 amplification → Trastuzumab deruxtecan
• NTRK fusion → Larotrectinib or entrectinib
• RET fusion → Selpercatinib

Step 3: If no actionable alteration and second-line was not FOLFOX/FOLFIRI:

• FOLFOX (preferred based on ABC-06 trial data)
• FOLFIRI (alternative)
• NalIRI+5FU (alternative)

Step 4: If all standard options exhausted:

• Clinical trial enrollment
• Best supportive care

Critical Pitfall to Avoid

Do not assume checkpoint inhibitor re-challenge is reasonable simply because the patient had a biomarker-selected indication (MSI-H/dMMR/TMB-H) initially. Progression on pembrolizumab indicates acquired resistance, and repeating the same mechanism is futile. The one exception would be switching from anti-PD-L1 to anti-PD-1 (or vice versa) in highly selected cases, but even this lacks robust evidence in cholangiocarcinoma.

Quality of Life Considerations

Third-line chemotherapy (FOLFOX/FOLFIRI) carries significant toxicity burden:

• Neuropathy (oxaliplatin)
• Diarrhea (irinotecan)
• Myelosuppression (both)

Performance status must be ECOG 0-1 to justify aggressive third-line therapy. If ECOG ≥2, best supportive care with palliative intent is more appropriate to preserve quality of life without meaningful survival benefit.